To all of the EJ bashers

[billsfan1959]

Et tu....what? You know how I approach problems: with unfettered logic, PROPER analytical methodology, and a full command of the English dictionary.

 

When I say EJ is a Rookie, that's because by all reasonable definitions, logically, statistically, and in NFL football English: he is.

 

I am quite tired of the "special rules for EJ" that I see on this board, from both sides.

 

Much could be gained by everyone here: watching another football game, paying attention to the QBs, and seeing, objectively, how often they FAIL/WIN, then, comparing what you see, once again, objectively, to EJ. Now that I am once again shackled by the taskmaster that is fantasy football(3 teams no less), I'm back to having to do precisely that.

 

Let me tell you all: the QB play I've seen around this league, especially from the "stars", so far, does NOT have the variance it once did. This sure as hell is NOT the early 90s. There's plenty of FAIL, especially from the "stars", doubly so for the "young stars"(um, RG3, Kapernick, Wilson, anyone?), to go around. Imagine if Tannehill was our QB....you don't have to: go read the boards of other teams around the league. Hardly anyone is happy with their QB play.

 

 

Then ask yourselves this question: why was ESPN's Sunday Top Ten Plays focused on "one-handed catches"?(Spiller's run came in at #10)

Answer: because exactly NONE of those balls were thrown accurately, by the EJ-hater side's "special rules for EJ" standard.

Spot on...

[Kelly the Dog]

 

 

Was Aaron Rodgers a rookie when he started his first game for Green Bay? Three years after he was drafted?

Apparently. EJ is not a rookie. His second year in the play book, two full years of OTAs, almost two years of training camp, an extra few months of off season work, almost two full preseasons, even time spent on the sidelines when he was hurt last year for six games is an advantage.

 

He has less than 16 starts that a rookie may have. That is the only thing that should be considered. It's a big one. But if you're going to say look at how good he is after 16 games, he would have an ENORMOUS advantage over all other rookies 16 games and it wouldn't be fair to them.

[Thurmal34]

 

Yep. Grabbed the wrist and pulled it down. No ball contact.

[papazoid]

Last season, E.J. Manuel was the second-worst quarterback to drop back to pass more than 100 times according to our NEP data, and his future was already looking a bit grim, despite only being a rookie.

 

But as of this moment, he sits in the 10th position in our Passing NEP rankings, an obvious step forward for the former first-round pick.

 

The Bills take on the Chargers this week, who were seventh from the bottom in passing defense last year, and have started this year as the 12th worst team in our NEP ranks. Manuel has looked good, and is developing chemistry with Sammy Watkins, which will help him post a solid day for fantasy owners against the San Diego.

 

http://www.numberfire.com/nfl/news/3176/week-3-fantasy-football-quarterback-streaming-ej-manuel-s-time-to-shine

[BillsBytheBay]

Two weeks of very tasty crow.

Continuing to shut up.

 

Signed, an EJ hater.

[YoloinOhio]

Last season, E.J. Manuel was the second-worst quarterback to drop back to pass more than 100 times according to our NEP data, and his future was already looking a bit grim, despite only being a rookie.

 

But as of this moment, he sits in the 10th position in our Passing NEP rankings, an obvious step forward for the former first-round pick.

 

The Bills take on the Chargers this week, who were seventh from the bottom in passing defense last year, and have started this year as the 12th worst team in our NEP ranks. Manuel has looked good, and is developing chemistry with Sammy Watkins, which will help him post a solid day for fantasy owners against the San Diego.

 

http://www.numberfire.com/nfl/news/3176/week-3-fantasy-football-quarterback-streaming-ej-manuel-s-time-to-shine

the CBs were indeed a weak link on the chargers last season. They were starting Crezdon Butler at one point, a player the Bills has cut in TC. They had a very poor FA signing in Derek Cox who they have big money to and ended up benching at the end of the year and then cutting in the off-season. But this year it is different. They signed Brandon Flowers, a pro bowler from KC. They drafted Jason a Verrett in the 1st rd, who has been outstanding early on. I think that unit should be much improved. Flowers has a groin issue and was out for the Seattle game, unsure of status for Sunday.

[YoloinOhio]

Michael Purinton ‏@NickelCityBills 49s

New graphs and stuff! MT @BuffaloWins: In case you missed...EJ Manuel's Development in the First Two Weeks - http://buffalowins.com/buffalo-bills/beyond-the-stats-by-nickelcitybills-ej-manuels-development-in-the-first-two-weeks.html …

[Solomon Grundy]

There are a lot of folks nationally that are still not convinced in EJ. Matt Miller, Charley Casserly, the "draftniks" over at Walter Football. They have the Bills drafting a QB in rd 2 stating that EJ will be a bust. I say we tell them it's CHICKEN instead of CROW maybe they'll eat it.

[papazoid]

funny how 10 games of below average play from last season (ave rank 27th) is not a big enough body of work....

 

but 2 games of slightly above average play from this season (ave rank 15th) is proof positive.

[John from Riverside]

funny how 10 games of below average play from last season (ave rank 27th) is not a big enough body of work....

 

but 2 games of slightly above average play from this season (ave rank 15th) is proof positive.

 

So there is literally identifiable proof of improvement.....

 

Thanks for pointing that out

[OCinBuffalo]

Say what? By dictionary definition, for a sportsperson/athlete, he most definitely isn't a rookie.

Horsecrap. 16 games started is the only real measurement we can use. What's the alternative? Running Madden simulations for the 6 games EJ didn't play last year? (You know it wouldn't surprise me...) Letting yet another "sports analytics" clown fail at doing my job, causing me to have to post, yet again, another long refutation of their idiot methodolgy?

 

(Past examples: "Road Wins Against Playoff Teams" , "Yards Per Attempt" )

Was Aaron Rodgers a rookie when he started his first game for Green Bay? Three years after he was drafted?

Ridiculous comparison. Was Aaron Rodgers starting games those 3 years? No.

 

In fact, you're proving my point for me: any analyis of a drafted-->starting QB is relatively pointless, even after 16 games, because of the plethora of variables that cannot be accounted for in a SANE manner.

 

Why is it that for last 20 years, we have always reserved "bust" status until 3 years after a player was drated, and rarely declared someone a bust until they demonstrate absolute suckitude consistently? Because that's the rational approach.

 

Suddenly, due to a few outliers like Kapernick and Wilson, we are supposed to throw away everything we know about the college-->NFL transition? Who the F gave that order, and why the hell are we following it? Again, there are FAR too many variables(like, um, the strength of both SF and SEA defenses?) to pretend we know a damn thing about Kapernick or Wilson or EJ so far.

 

Question: How much is EJ been benefitting from our D/ST(like Kap and Wilson have) thus far vs. how much is due to EJ himself?

No one has any idea until we get more data.

 

I'll use our current game plan over the last 2 games as an example: Would this be the gameplan GB would have used Aaron Rodgers in his post-Farve start? Of course not. So, does Aaron Rodgers first 16 games have much chance of telling us anything in relation to EJ's first 16 games? Not a chance in hell.

 

Using statistical analysis requires precision, it also requires knowing WTF you are doing. I've seen very little precision, all sorts of confidence bias, and hardly anyone demonstrating proficiency in this new "field" of "sports analytics" thus far. Football Outsiders and PFT are the only people "doing it right" and even they have flaws in their methods. The difference? They are competent enough to know about these flaws, they admit them, and they are trying to fix them. QBR is flawed as well, but, ESPN has decided that this is the best that can be done given the resources they are willing to put into it. Fine. At least it has a reasonable methodology.

 

In comparison, I hear morons talking about YPA, without even demonstrating the slightest correlation, never mind finding a propensity %, to winning/making the playoffs, not even a relationship to offensive proficiency/efficiency. Why? Because they haven't even bothered to set a F'ing baseline for these things.

 

Yards Per Attempt sits there like a burning pile of schit, and we have moronic "shamans" howling at the moon and dancing around it, and getting the same exact results in terms of predicting future results.

Last season, E.J. Manuel was the second-worst quarterback to drop back to pass more than 100 times according to our NEP data, and his future was already looking a bit grim, despite only being a rookie.

 

But as of this moment, he sits in the 10th position in our Passing NEP rankings, an obvious step forward for the former first-round pick.

 

The Bills take on the Chargers this week, who were seventh from the bottom in passing defense last year, and have started this year as the 12th worst team in our NEP ranks. Manuel has looked good, and is developing chemistry with Sammy Watkins, which will help him post a solid day for fantasy owners against the San Diego.

 

http://www.numberfir...s-time-to-shine

Here's another interesting, QBR-like approach to rating players. Methodology here: http://www.numberfir.../info/glossary/ First let me say, this is at least a competent effort.

 

The problem is, just like with QBR, and to quote the method explanation itself:

That's a ton of variables[Finally somebody is starting to get it], but luckily, numberFire has data from the past dozen years of every single play, so most situations have come up at least once.

 

The "at least once" part is the issue. Using historical data in this manner doesn't account for deltas in team defensive schemes, or offensive schemes, nor does it account for individual opposing player's strengths weaknesses, nor does it account for time of game(or perhaps it does, but this doesn't say.)

 

This is why PFT's rating schema is superior: it compares player v. player, every play, all game. (But it has other problems that I won't get into here)

 

While it's fair to assume that "rare" plays, of the "at least once" variety have a reasonable chance of remaining "rare", how does it account for something like Converting a 4th Down and 32 with an inside handoff? That's an pretty rare/large achievement, and either indicative of one hell of a RB, or O line, or, indicative of a terrible D. Or, maybe it's merely a scheme thing? Time of game? Score? While, throwing for 6 yards on a slant, on 2nd/3rd and 5, all over the field, any time of game, happens all the time.

 

Thus, you're bound to have a bias in "expected successful outcome achieved" in "most common situation". You're going to see the pattern you are looking for(otherwised known as confidence bias) in the most common situations, precisely because they are so common, and precisely because teams use "common" approaches to solving "common" problems. Almost every QB in the league is capable of converting that slant, and they do, lots.

 

How therefore does any of this distinguish one QB from another? It over-credits QBs for making that play, and over-punishes QBs for not making it.

 

Now, consider the fact that 12 years ago, nobody was running the read/option, and there were very few true "running QBs", therefore, no LB had to account for the QB(except for a very few) taking off. Now, many teams are designing runs for QBs(The Dolphins designed 2 runs for Ryan F'ing Tannehill last game). Thus, the run action is making the passing game easier for those QBs who bring that threat to the game. (Ahem, look at Wilson, Kapernick, and to a lesser degree, EJ)

 

This is where QBR-like approaches fail. It's the threat of the run that they simply do not take into account. They try to measure and then weight a QB run as more equivalent to the pass, but only AFTER the fact.

 

Passing will, in the aggregate, be better for all QBs who can demonstrate both a competent pass and run threat, especially for teams that gameplan that very thing, thus causing defesnes to respond with appropriate gamplans. But, data from 12-5 years ago doesn't take that into account, does it? EDIT: At the very least, you've got one skew here. Maybe more than a few.

 

In closing, this methodology is well thought out, and is a responsible, competent effot. But, clearly, so are my criticisms of it.

 

The bottom line: proper weighting of the raw data, BEFORE it's put into the algorithms, is required to make this method more accurate. Either that, or somebody simply needs to show me the alogrithms, or, somebody simply needs to tell me that the alogrithms have taken everything I've said into account....and make me believe it(tough gig).

Edited September 20, 2014 by OCinBuffalo

[Dibs]

Horsecrap. 16 games started is the only real measurement we can use. What's the alternative? Running Madden simulations for the 6 games EJ didn't play last year? (You know it wouldn't surprise me...) Letting yet another "sports analytics" clown fail at doing my job, causing me to have to post, yet again, another long refutation of their idiot methodolgy?

 

(Past examples: "Road Wins Against Playoff Teams" , "Yards Per Attempt" )

........

 

The only reason I initially responded was because you posted these 2 points...

....You know how I approach problems: with unfettered logic, PROPER analytical methodology, and a full command of the English dictionary.

....

......

When I say EJ is a Rookie, that's because by all reasonable definitions, logically, statistically, and in NFL football English: he is.

.....

 

The English dictionary definition is....

http://dictionary.reference.com/browse/rookie?s=t

1. an athlete playing his or her first season as a member of a professional sports team:

 

 

I think that you are a very intelligent person and I truly appreciate and value your posting.......but if you are going to act obnoxiously superior I will take the time to call you out on mistakes.

(btw, I agree with your premise and argument, but you called up the wrong definition).

[OCinBuffalo]

The only reason I initially responded was because you posted these 2 points...

 

 

 

The English dictionary definition is....

http://dictionary.re...owse/rookie?s=t

1. an athlete playing his or her first season as a member of a professional sports team:

 

 

I think that you are a very intelligent person and I truly appreciate and value your posting.......but if you are going to act obnoxiously superior I will take the time to call you out on mistakes.

(btw, I agree with your premise and argument, but you called up the wrong definition).

What mistake?

 

Again, I ask: what's the alternative to calling EJ a rookie until he completes 16 NFL games? Again I ask: 6 madden simulations? Listening to Mike Schopp tell us what would have happened, given EJ's Yards Per Attempt?

 

You have no logical alternative. Therefore, there is no error in what I am saying.

 

You're the one that bolded me saying "NFL football English" above. Why did you do that? I could have just said "English". But, I said "NFL football English", didn't I?

 

Think that was a mistake?

 

EDIT: Oh, and I am certainly not acting. When it comes to what is currently being called "analytics"(merely the lastest marketing buzzword for what has been one of my areas of expertise for 20 years)? I've been superior, for quite a while. And, when the word for "analytics", or any of the other things I do...changes yet again? I'll remain: superior.

 

I'll make you a scary prediction:

sooner or later, somebody is going to realize the "value" in applying what some other superior people around the world, and I have been working on for the last 10 years or so, known as "real time analytics"....to the NFL. Right now the "sports analytics experts" are only at doing what I was doing ~15 years ago(ouch, now I feel old).

 

IF "real time NFL analytics" were ever to be implemented by the same media jokers/amatuers that are f'ing about with the old stuff today, it's quite possible that a black hole of idiocy would be created, that none of us could ever escape.

 

....and that, is why I write these posts. Think: nothing but Dan Dierdorf absurd analysis, or, perhaps more likely, Joe Buck/Joe Theisman "praise whoever is winning, right now" analysis and/or "praise whoever the analytics guy tells me to....until he tells me not to...I don't care if I just contradicted what I was saying 3 minutes ago: IT'S REAL TIME ANALYTICS!" analysis 24/7.

 

Nightmare, isn't it?

Edited September 20, 2014 by OCinBuffalo

[Dibs]

What mistake?

 

Again, I ask: what's the alternative to calling EJ a rookie until he completes 16 NFL games? Again I ask: 6 madden simulations? Listening to Mike Schopp tell us what would have happened, given EJ's Yards Per Attempt?

 

You have no logical alternative. Therefore, there is no error in what I am saying.

 

You're the one that bolded me saying "NFL football English" above. Why did you do that?

 

I could have just said "English". But, I said "NFL football English", didn't I?

 

Think that was a mistake?

 

Had you just said "English", you could have gotten away with the other definitions(raw recruit, a novice)......but you specifically said "NFL football English". In NFL football terminology(and every sport) a rookie is an athlete playing in their first season. By dictionary definition, EJ is not a rookie.

 

I reiterate that I have no problem whatsoever with your argument. EJ didn't play a full rookie season, therefore that should be factored into ones analysis when determining his progress etc, etc.

 

I merely had a problem with you saying(in a superior manner) that you had "a full command of the English dictionary" and then directly followed that up with "...EJ is a Rookie....in NFL football English".

 

 

For somebody who avows to have a full command of the English dictionary, it is ironic that you cannot understand that you made a simple error in regards to the definition that you used.

Edited September 20, 2014 by Dibs

[Thurmal34]

funny how 10 games of below average play from last season (ave rank 27th) is not a big enough body of work....

 

but 2 games of slightly above average play from this season (ave rank 15th) is proof positive.

 

Don't we all want the QB of the Bills to improve? We all know Manuel isn't Luck, but if he continues to improve he'll be just fine for this team.

 

[HOUSE]

Bandwagons went out in the 1920's

 

Is this grampa Wiggy ? I miss you

[papazoid]

 

 

So there is literally identifiable proof of improvement.....

 

Thanks for pointing that out

 

going forward, if EJ has two below average games in a row, will that be identifiable proof of regressing ?

[Thurmal34]

going forward, if EJ has two below average games in a row, will that be identifiable proof of regressing ?

 

Regressing from the improvements he has made this year or regressing from the below average standard he set last year?

 

Posit for a moment, if you will, that EJ's next two games are below average. Would you consider that to be identifiable proof of regression from this year's standard that he has set in games 1 and 2, or some other papazoidian standard that you arbitrarily set? What is the benchmark for regression?

Edited September 20, 2014 by Thurmal34

[Maury Ballstein]

Keep rooting against him papazoid. Sorry this 2-0 start is so painful.

[Solomon Grundy]

Technically, it does say "playing".